The Open Source reference architecture for 3D Digital Twins in AgTech, Industry, and Environment.

Nekazari OS (NKZ) is a highly modular, GitOps-driven platform designed to bridge the gap between physical assets (Edge) and their digital representations (Cloud). Built strictly on FIWARE NGSI-LD and Smart Data Models, it provides unparalleled interoperability for massive satellite imagery processing, IoT telemetry, and autonomous agricultural robotics.

NKZ OS follows a decoupled architecture to prevent vendor lock-in and allow independent scaling of components.

• 🧠 nkz: The Core backend. Contains the Context Broker integrations, FastAPI Gateway, TimescaleDB persistence, Keycloak Auth, and the CEP visual rules engine.
• 🌐 web: The SSG portal for our Docs-as-Code automated documentation.
• 📦 nkz-module-template: The canonical starter kit for third-party developers to create platform extensions with pre-configured CI/CD and Docker scaffolding.

NKZ OS relies on an IIFE (Immediately Invoked Function Expression) Runtime Script Injection architecture. This allows third-party developers, academic researchers, and hardware manufacturers to build independent, domain-specific modules (e.g., LiDAR point clouds, isoBUS telemetry, Crop Health via Sentinel-2) that dynamically plug into the core CesiumJS 3D engine without tight coupling or redeploying the core platform.

We provide Hardware Reference Designs for field edge devices. Whether you are building custom MQTT dataloggers or designing autonomous agricultural rovers (ROS2/Zenoh), our reference architectures ensure your hardware is natively compliant with the NKZ Digital Twin environment right out of the box.

👉 Check the Edge Integration Docs for BOMs, isoBUS wiring, and payload schemas.

NKZ is built to be the foundational infrastructure for AgTech research. We actively welcome contributions from university talent (Computer Science, Telecommunications, Agronomy).

• Don't waste time building backends for your TFG/TFM. Fork the nkz-module-template, focus on your algorithm (AI, Vision, Robotics), and inject it into our 3D environment.
• Look for issues tagged with good first issue or tfg-candidate across our repositories.

For public institutions handling Copernicus data or managing large-scale agronomic knowledge transfer, the NKZ architecture handles terabytes of TIFs via MinIO and rio-tiler without cluster degradation. The NKZ Enterprise deployment model offers SLA-backed operations, dedicated GitOps pipelines (ArgoCD), and custom legacy ERP integrations (Odoo).

All technical documentation, K8s deployment guides, and API specifications are automatically compiled from our source code:

👉 Read the Official Documentation at nkz-os.org

📫 Contact the maintainers: nkz@nkz-os.org